OpenAI Wants Your Medical Records Now
Handing over your cholesterol levels and daily step count to an AI chatbot sounds like a plot from a dystopian sci-fi novel. Yet here we are. OpenAI just rolled out ChatGPT Health to all US users, marking one of its boldest consumer pushes yet. You can now feed the bot your real-time data from Apple Health, MyFitnessPal, and Function Health.
It's live right now. If you're in the United States, you can link your accounts with a few taps. So, why does Sam Altman want to know how many calories were in your lunch? Simple. The race to dominate consumer AI isn't just about answering trivia or writing code anymore. It's about becoming indispensable.
How ChatGPT Health Actually Works
Here's what most coverage misses: this isn't just a gimmick to show you pretty charts. When you plug in your data, ChatGPT starts cross-referencing your lifestyle choices with actual biological metrics. MyFitnessPal sends over your macronutrients. Apple Health brings sleep trends and heart rate variability. Function Health tosses in comprehensive blood panels and biomarker results.
Put them together, and you get a conversational engine that understands why you felt sluggish on Tuesday. It sees you slept four hours, ate 300 grams of carbohydrates, and had elevated cortisol levels on your last lab test. When stacked up in direct matchups like ChatGPT vs Gemini, OpenAI is clearly doubling down on deep integrations over generic web search tools.
And people will use it. Millions of them. The reality is that US healthcare is expensive, bureaucratic, and frustratingly slow. Waiting three weeks for a doctor to spend four minutes explaining your lipid panel is a terrible experience. Getting an instant, plain-English breakdown from an AI feels useful, even if it comes with disclaimers printed in tiny font at the bottom of the screen.
Privacy Hazards and the Moat Strategy
That said, trusting an AI company with your medical diagnostics raises massive red flags. We've watched models hallucinate factual errors before. When a bot misinterprets a coding script, you waste an hour debugging. When it misinterprets a lab result, the consequences are vastly different. OpenAI promises that health data is isolated and won't be used to train its base models without explicit permission. But given how OpenAI's AI spending spree has ballooned to $750B, the commercial pressure to find new monetization avenues is immense.
So why take the medical risk? Because lock-in is the ultimate moat.
Right now, switching costs in consumer tech are low. Users bounce between options depending on which model dropped an update this week. When users compare ChatGPT vs Claude for writing tasks or ChatGPT vs Perplexity